Complementing habitat distribution model with land use land cover for conservation of the rare and threatened tree Magnolia punduana Hk. f & Th. in northeast India.

Predictive distribution models are widely used in species conservation planning. However, identifying specific sites for undertaking conservation action such as reintroduction pose a practical challenge for practitioners. We demonstrate that the potential habitat distribution models intersected with existing land use and land cover (LULC) map helped in identifying the specific conservation areas with greater confidence. We used a case of Magnolia punduana Hk. f & Th. (Magnoliaceae)—a threatened and rare tree species distributed in the Khasi, Jaintia and Garo Hills of northeast India at an elevation range of 800–1600 m asl. We modeled the distribution of potential habitats of the species using maximum entropy (Maxent) software, moderate resolution imaging spectroradiometer (MODIS) imageries, and ASTER-based digital elevation data. Amongst the environmental variables, elevation and the EVI for the months of May, June and July had a collective contribution of > 70%, indicating their dominant role in determining the potential habitat. The predicted potential habitat of the species covered ~ 2.88% area of the state in the southern part of the Khasi and Jaintia hills with a few patches in Garo hills. However, intersection of the potential distribution area map with the existing forest cover map of the study area showed ~ 3853 ha of the state, has high suitability. The ground-truthing of the high suitability regions revealed that they are similar to the species' original habitat and may support conservation-related activities such as habitat protection, restoration, and species (re)introduction. [ABSTRACT FROM AUTHOR]